🎮

Jensen Huang

Original nameJensen Huang

Co-Founder and CEO of NVIDIA, the Key Figure Who Turned the Gaming Graphics Card Into an AI Compute Engine

Those Who Turned Invention Into Industry · Industry-Builders

GPU Parallel ComputingThe CUDA PlatformAI Compute Infrastructure

Who they are

Jensen Huang (1963– ) was born in Taiwan, emigrated to the US in childhood, and in 1993 co-founded NVIDIA with two other engineers, at first focusing on designing high-performance graphics processing chips (GPUs) for video games. In 2006 the company launched the "CUDA" platform, letting developers apply the large-scale parallel computing power of GPU chips, originally designed for graphics rendering, in a general-purpose way to scientific computing and data-processing tasks beyond graphics — a strategic decision that did not win wide market attention at the time, yet into which NVIDIA continued to pour R&D resources for years. After the rise of deep learning, researchers gradually found that the large-scale matrix parallel operations required by neural-network training fit highly with the computing architecture GPU chips were originally optimized for graphics rendering, and in 2012 "AlexNet" was trained precisely on two NVIDIA GPUs. GPUs thereafter swiftly replaced traditional processors to become the indispensable core compute infrastructure of global deep-learning and generative-AI model training, and NVIDIA, riding this wave of explosive AI compute demand, saw its market value leap for a time to the front rank of global technology firms. For this "ten years to sharpen one sword" long-term strategic vision, Huang is widely seen as one of the most important infrastructure drivers behind this wave of the AI industry.

Primary sourcesNVIDIA historical financial reports and CUDA platform technical documents

Key stories

A Bet That Took Over a Decade to Pay Off

When NVIDIA launched the CUDA platform in 2006, the idea of general-purpose GPU computing was of exceedingly limited commercial value and market demand at the time, yet the company continued to pour large R&D resources into perfecting the platform; this long-term strategic bet, quite risky in outsiders’ eyes, at last saw a true market explosion nearly a decade later, after the rise of deep learning, when GPU parallel computing power happened to fit the needs of neural-network training.

The Gaming Graphics Card Becomes AI Infrastructure

NVIDIA’s original core goal in designing GPU chips was only to render video-game images more smoothly and realistically, but the architectural feature of the GPU chip of "handling many simple parallel operations at once" happened to fit highly with the large-scale matrix operations required by deep-neural-network training, and this chip technology originally for the entertainment industry at last accidentally became the most core compute infrastructure driving the whole wave of the AI industry.

Relationships

Disciple

🤖Sam Altman　OpenAI’s later large-scale training and deployment of the GPT series of models relied heavily on the compute infrastructure NVIDIA GPUs provided

Echoes today

Below are how modern works borrow or reinterpret this name or story — not the original material. The two differ, so keep them apart.

The signature leather-jacket imageHuang almost always appears in public in a black leather jacket, a highly recognizable personal image that thereafter became a visual symbol recurring in tech-media coverage of NVIDIA and the topic of AI compute.

Appears in

The Explosion of Deep Learning and Generative AIJensen Huang’s Decade-Long GPU Bet

← Back to Figures